3D Visual Fatigue Determination via Depth-of-Field Analysis

The patent divides the image into multiple regions of interest (such as foreground objects, background, and different depth zones) and calculates depth-of-field values separately for each region. This segmentation approach reduces the overall computational complexity by processing smaller subsets of pixels independently, while still achieving comprehensive visual fatigue assessment across the entire image.

The patent applies different depth-of-field calculation methods and weighting factors to different spatial regions of the image based on their importance. Critical regions (such as main subjects or areas with high visual impact) are analyzed with higher precision and greater weight, while less important regions use simplified calculations. This local quality approach optimizes the balance between measurement accuracy and computational effort.

The patent pre-calculates and stores depth maps, parallax information, and region-of-interest masks for 3D images and videos before performing visual fatigue assessment. These preliminary computations are saved as metadata or intermediate results, allowing the actual visual fatigue calculation to reuse pre-computed data and avoid redundant calculations, thereby reducing processing time.

The patent implements incremental updating of depth-of-field values and visual fatigue metrics when processing video sequences. Instead of recalculating all depth values from scratch for each frame, the system maintains and updates only the changed regions or uses motion compensation to predict depth values for consecutive frames, ensuring continuous assessment with minimal computational overhead.